Visualization

Pulling data from API (shout-out to Chris)

Chris provided this part of code. The idea is to pull all the session ids connected to the username NYCEJA, and saved in an R object df

real <- c("http://aircasting.org/api/realtime/sessions.json?page=0&page_size=500&q[measurements]=true&q[time_from]=0&q[time_to]=2552648500&q[usernames]=NYCEJA")

tt <- jsonlite::fromJSON(real)
t <- tt$streams$'AirBeam2-PM2.5'
t <- data.table(t)
ID <- t[!is.na(t$id)]$id

name <- data.frame(tt$title,t$id)
name <- name[!is.na(name$t.id),]
colnames(name) <- c("title", "id")

dt <- list()
for (i in 1:length(ID)) {
  sess <- paste0("http://aircasting.org/api/realtime/stream_measurements.json/?end_date=2281550369000&start_date=0&stream_ids[]=",ID[i])
  s1 <- jsonlite::fromJSON(sess)
  s1 <- data.table(s1)
  s1$ID <- ID[i]
  dt[[i]] <- data.table(s1)
}

library(plyr)

## -------------------------------------------------------------------------

## You have loaded plyr after dplyr - this is likely to cause problems.
## If you need functions from both plyr and dplyr, please load plyr first, then dplyr:
## library(plyr); library(dplyr)

## -------------------------------------------------------------------------

## 
## Attaching package: 'plyr'

## The following objects are masked from 'package:plotly':
## 
##     arrange, mutate, rename, summarise

## The following object is masked from 'package:lubridate':
## 
##     here

## The following objects are masked from 'package:dplyr':
## 
##     arrange, count, desc, failwith, id, mutate, rename, summarise,
##     summarize

## The following object is masked from 'package:purrr':
## 
##     compact

df <- ldply(dt, data.frame)

check df. This data contains measurement from 23 locations (represented by 23 IDs), generated every minute from 2018-6 to 2019-3. Note that the earliest date is in 2000 which does not make sense. This is more likely to be 2019, so I will change all 2000 to 2019.

# check data
head(df) %>% 
  knitr::kable()

Data cleaning

A bit data cleaning and manipulation

air_data <- df %>% 
  as_tibble() %>% 
  mutate(
    time = str_replace(time, 'T', ' ')
  ) %>% 
  mutate(
    time = str_replace(time, 'Z', '')
  ) %>% 
  mutate(
    time = str_replace(time, '2000', '2019')
  ) %>% 
  mutate(
    time = str_replace(time, '1999', '2018')
  ) %>% 
  mutate(
    time = ymd_hms(time),
    ID = as.character(ID)
  ) %>% 
  dplyr::rename(
    station_id = ID
  ) %>% 
  filter(
    !is.na(time)
  )

Visualization

location <- df %>% 
  distinct(latitude, longitude) %>% 
  slice(-9, -20) %>% 
  mutate(
    name = str_c(round(latitude, digits = 2), round(longitude, digits = 2), sep = ', ')
  )

leaflet() %>% 
  addTiles() %>% 
  addCircleMarkers(
    data = location,
    lat = ~latitude, lng = ~longitude,
    color = 'green',
    label = ~htmlEscape(name)
  )

location_3 <- air_data %>% 
  filter(
     station_id == '246560'
  ) %>% 
  mutate(
    measure_date = as.Date(time),
    measure_hour = hour(time)
  )

loc_3 <- location_3 %>% 
  mutate(
    timepoint = str_c(measure_date, measure_hour, sep = ' ')
  ) %>% 
  arrange(measure_date, measure_hour) %>% 
  mutate(
    row_num = rownames(.)) %>% 
  mutate(
    row_num = as.numeric(row_num)
  ) %>% 
  ungroup() %>% 
  ggplot(aes(x = row_num, y = value)) + geom_line() + geom_smooth()

ggplotly(loc_3)

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'